Evening,
Have a strange issue.
When we restart SQL Server we are unable to connect for almost 2 hours afterwards.
Connecting to SQL does not fail it just hangs.
If I open the active SQL errorlog in notepad I can see SQL starting up, bringing the other system DBs up saying it can now accept connections and then it says it is clearing the TempDB database then almost 2 hours later it says starting up TempDB database.
I am aware of the process in having TempDB rebuilt every time SQL gets restarted so I am figuring there is something that is slowing down the dropping of the old TempDB files and recreating the new one.
Originally we had TempDB being created on a SAN volume so I thought while the SAN guys checked out their environment I would redirect TempDB to rebuild on a local disc.
Same issue ... very long time between TempDB clearing and starting.
Below is a rundown of the info I sent my SysOps and SysEng group:
=========
Last couple of reboots we did not wait the full 2 hours for the SQL Server restart to finish up (which is course is way way too long).
I looked back yesterday around 5:19 when we rebooted it and left for the day … we can see the same issue … TempDB is cleared at 5:19 and is not rebuilt and started until 7:04.
2013-07-30 17:19:25.21 spid25s Clearing TempDB database.
2013-07-30 17:19:28.50 spid28s CHECKDB for database 'Admin' finished without errors on 2013-07-20 06:10:07.447 (local time). This is an informational message only; no user action is required.
2013-07-30 17:19:28.50 spid28s Recovery completed for database Admin (database ID 5) in 9 second(s) (analysis 9060 ms, redo 0 ms, undo 75 ms.) This is an informational message only. No user action is required.
2013-07-30 19:04:06.43 spid25s Starting up database 'tempdb'.
I went back to our oldest SQL event log file which was created on the 27th around 8:20p (SQL is set to only keep 7 archives of the SQL event logs after each restart plus the active log for a total of 8 versions) and I found this event:
2013-07-27 20:22:34.21 spid25s Clearing tempdb database.
2013-07-27 21:17:31.92 spid25s Starting up database 'tempdb'.
4 days ago it still had a large lag time between dropping and recreating the TempDB files but it only lasted a little less than an hour.
===========
So would any of you fine Gurus have any insight as to what I may be encountering.
As a couple of tidbit here ...
We are running the instance on a Win 2K8 R2 server w/ SQL 2K8 R2 SP1.
Using port 11005 for SQL traffic and default 1434 for DAC.
Due to business requirement we only have TCPIP & Share Memory enabled.
I am not seeing anything out of the norm in the Win App and Sys event logs.